Liquid crystal display (LCD) technology has progressed rapidly in recent years, and has become an increasingly common option for display systems, currently making up the largest portion of the flat panel display market. This market dominance is expected to continue into the future. The superior characteristics of liquid crystal displays with regard to weight, power, and geometry in image visualization, have enabled them to compete in fields historically dominated by Cathode Ray Tube (CRT) technology, such as high definition television systems, desktop computers, projection equipment, and large information boards. As the cost of LCD systems continues to fall, it is predicted that they will eventually take over the market for traditional CRT applications.
The biggest disadvantages of current CRT systems are their geometrically bulky size and weight, as well as their high power consumption. These disadvantages are clearly evident when comparing the features of CRT and LCD projection displays with similar characteristics. In general, projection display systems offer several additional advantages over CRT systems. First, projection display systems offer the possibility of using large screens for group viewing with the ability to easily change the image size and position. Second, projection display systems offer high performance, and the ability to accept image data input from a variety of devices such as computers, television broadcasts, and satellite systems. Virtually any type of video input can be projected through such a system. The application of LCD's to projection systems has further attractive features such as high brightness, high resolution, and easy maintenance. LCD front projection displays provide higher resolution and brightness than comparable CRT-based systems. In comparison with CRTs, installation of LCD projection systems is easy and their viewing angles are generally much wider. Most front projection LCD display systems are compatible with personal computers and can operate with video signals from television systems. LCD front projectors are easily adapted for applications such as home theaters.
Typically, LCD projection systems include one or more small LCD panels, usually ranging from 1 to 5 inches in diagonal, a series of dichroic mirrors or filters, and a series of projection lenses. Commonly, three LCD panel systems are used, where one or more dichroic mirrors divide white light coming from a light source, into the three primary colors of red, green, and blue (RGB). The dichroic mirrors direct each of the RGB components toward a separate LCD panel. The corresponding LCD panel modulates each of the RGB components of the light according to data from an input device. Output dichroic mirrors synthesize the modulated RGB light components and project the image onto a viewing screen.
To enhance the luminance of the liquid crystal projection panels, reflective LCD pixels are used. These systems, sometimes referred to as Liquid Crystal on Silicon micro-display (LCOS), utilize a large array of image pixels to achieve a high resolution output of the input image data. Each pixel of the display includes a liquid crystal layer sandwiched between a transparent electrode and a reflective pixel electrode. Typically, the transparent electrode (sometimes called the ITO layer) is common to the entire display while the reflective pixel electrode is operative to an individual image pixel. A storage element, or another type of memory cell, is located beneath each of the pixels and is operative to direct a voltage on the pixel electrode. By controlling the voltage difference between the common transparent electrode and each of the reflective pixel electrodes, the optical characteristics of the liquid crystal can be controlled according to the image data being supplied. Generally, the optical characteristics of liquid crystal materials are responsive to an applied voltage. The storage element can be either an analog or a digital storage element. More and more often, digital storage elements, in the form of static memory are being used for this purpose.
The liquid crystal layer modifies the polarization state of light that passes through it. In digital systems utilizing nematic liquid crystals, the extent of the modification to the state of polarization of incident light depends on the root-mean-square (RMS) voltage that is applied across the liquid crystal layer. The intensity of the reflected light depends therefore on the proportion of reflected light that is orthogonal to the polarization state of the incident light. (Sometimes referred to as “on state” light.) This value is in turn determined by the voltage being applied to the pixel electrode by the storage element, such being well known to those of ordinary skill in the art.
Therefore, by applying varying voltages to the liquid crystal material, the liquid crystal device can be configured to return varying amounts of “on state” light. When controlled by a digital storage element that can supply one of two possible instantaneous voltages to the pixel electrode, the liquid crystal material will respond in one of two principal ways, depending on the material. In the first instance, where the liquid crystal response time is much faster than changes to the drive waveform, the polarization state encoded into the reflected beam will closely follow the original drive waveform. In the second instance, where the liquid crystal response time is much slower than the changes to the drive waveform, the polarization state encoded into the reflected beam of light will follow the RMS of the applied voltages. In either instance the liquid crystal acts as a variable optical retarder, rotating some, all, or none of the incident polarized light, resulting in a varying intensity of the reflected beam of light once analyzed by a polarizing device. A human observer looking at the beams of light created by such devices will tend to average the intensities over a time scale of 15 to 30 milliseconds. Thus either modulation result can be resolved by human observers as gray scale images, provided the time frames for the different intensities are suitable short in duration. Finally, by varying the amount of time that the pixel is either “white” or “black,” the human eye will perceive a gray scale shading somewhere between totally white and totally black.
Gray scale modulation may be used in a display to permit the display of a full range of colors. As is well known in the art, a reasonably complete range of colors can be created by combining the primary colors (red, green and blue) in varying intensities. The total number of different colors that can be created are determined by the number of gray levels that are available in a given color generation system. The gamut of the colors that can be created is determined by the spectral composition of the individual primaries. Thus the generation of gray levels in a pixilated display is a critical element in the capability of such a system to generate realistic images.
Pulse-width-modulation (PWM) is a method of driving these types of digital circuits to create gray scale. In one type of PWM, varying gray scale levels are represented by multi-bit words (i.e. a binary number). These multi-bit words are converted into a series of pulses. The time averaged RMS voltage corresponds to a specific voltage necessary to maintain a desired gray scale.
Another method for creating gray scale is binary-weighted pulse-width-modulation, where the pulses are grouped to correspond to the bits of a binary gray scale value. The resolution of the gray scale can be improved by adding additional bits to the binary gray scale value. For example, if a 4 bit word is used, the time in which a gray scale value is written to each pixel (frame time) is divided into 15 intervals. This results in 16 possible gray scale values (24 possible values). An 8 bit binary gray scale value would result in 255 intervals and 256 possible gray scale values (28 possible values).
In addition to controlling the RMS voltage that is seen by the liquid crystal material in each of the display pixels, modulation schemes may be incorporated that control how the specific data is written to the display imager (as opposed to how each pixel reacts to the supply voltage). Liquid crystal imagers consist of a series of pixel rows, and known systems write data to the imager one row at a time, typically beginning with the top row of the imager and sequentially progressing through all of the rows in the display. For example, in a VGA display that has 480 rows of pixels, and 640 pixels per row, a known modulation scheme would write data to each of the pixels in the first (i.e. top) row, and then progress to the next row in line and write data to each of the pixels in that next row. This scheme repeats until all 480 rows have been written. The process then repeats from the first row, updating the data reflected in each pixel depending on the image that is to be displayed. Under this known scheme, an individual pixel value is changed once every n row write times, where n is the number of rows in the imager (e.g. 480 rows in a VGA system). With the current level of resolution that LCD displays are achieving (i.e. 2000.times.1000 pixels in a HDTV scheme), the amount of time that a pixel waits to be rewritten is drastically affected in these once through write schemes.
In known systems using pulse width modulation, a higher imager write frequency improves the modulation efficiency, since the data for each pixel can be updated more frequently. However, the time that each bit of data is displayed also needs to be controlled and thus higher frequency systems do not always solve the control problem. Furthermore, higher speed driving circuits are inevitably more expensive and draw more power from the system, factors that are undesirable in the design of such circuits.
Another way to improve the modulation efficiency is to lower the frame rate of the system. However, this solution will significantly aggravate flicker issues in the display, another undesirable effect. It is therefore desirable to increase the imager write frequency in a display without increasing the frequency of the driving circuit and without increasing the system power consumption.
Both digital and analog modulation schemes suffer from lateral field defects, where two adjacent display pixels, one at a high voltage and one at a low voltage, have a very high pixel-to-pixel (i.e. lateral) field strength. This lateral field strength is commonly on the order of 10 times the vertical field strength. Since the two adjacent pixels represent a black to white, or dark to light, transition, the lateral field, which highlights the transition, is not a strong visual artifact and ultimately distorts the image. Notably, the transition between the two adjacent pixels (the edge) will be enhanced and the image will not appear as clear.
In this situation, digital modulation schemes are even more severely constrained because gray levels in adjacent pixels can produce lateral field effects (pixel-to-pixel) that are high enough to overpower the desired vertical field effect (pixel-to-ITO). The vertical field effect is what ultimately determines what gray scale value is displayed through the pixel. For a digital modulation scheme that utilizes simple binary weighted pulse width modulation, objectionable lateral field contours (defects) occur, for example, where adjacent pixels are driven at the mid gray levels 7 f and 80 (100% pixel-to-pixel temporal intermodulation), at ¼ gray levels 3 f and 40 (50% pixel-to-pixel temporal intermodulation), and at ⅛ gray levels if and 20 (25% pixel-to-pixel temporal intermodulation). These represent instances where the data in adjacent pixels are out of phase to the degree indicated and where the inter-pixel modulation lines resulting from the lateral fields stand out in sharp contrast to the modulation levels of the two pixels. While thermometer based codes can ameliorate the digital-unique lateral field effects with an increased frequency and an increased number of time divisions (normally a 2.times. improvement for 2.times. increase in bandwidth), this also aggravates the modulation efficiency because there is a trade off with the lateral field defects. See Yang, et al. IBM Journal of Research and Development, Volume 42, Number 3/4, May/July 1998, pp. 405-407, the contents of which are incorporated herein by reference, for an additional description of lateral field effects and reverse-tilt disclinations in nematic liquid crystal displays.
The inherent characteristics of liquid crystal materials also affect the modulation efficiency of these displays. For instance, reverse twists (multi-second smoke trails) limit the use of imagers that are based on either analog or digital modulation techniques. Known digital modulation schemes are more demanding on the liquid crystal material for reverse twist tolerances because of the higher driving voltages for use with common drive schemes. This also results in a reduced modulation efficiency.
Additionally, both analog and digital modulation schemes can suffer from flicker effects due to the use of low-frequency ITO drive schemes. The flicker frequency equates to half of the ITO-inversion rate. While this can have a more drastic effect on analog systems, digital pulse width modulation schemes result in a non-linearity in the digital code to RMS voltage mapping. This can both help and hurt the electro-optical curve linearization.
The modulation efficiency in known digital systems is limited for several reasons. First, the pixel voltage (V1) is turned off (i.e. not modulating a full white value) during the period of time the imager is being written with the next portion of the binary weighted data. V1 is then pulsed for the time associated with the next portion of the binary weighted data. This process repeats to write each portion of the image data. The limited time frame during which the write function can take place limits the modulation efficiency.
Second, even though applying an overlap of array-write and liquid crystal voltage drive improves the modulation efficiency, increased thermometer decoding limits the overlapped write improvements. Lowering the frame rate (rather than the peak frequency) also improves the modulation efficiency, but can significantly aggravate display flicker issues.
Third, known methods of gray scale modulation are suboptimal. For gray scale modulation, known digital displays typically write every row or write the entire display and then sequence the display so that there are two storage registers for each pixel. The display writes the first register and strobes the data to bring forward the second register to display it on the pixel. Unfortunately, this approach creates a problem whereby for the least significant bit (LSB) or lowest gray scale value, the write time for the display may be longer than the duration of the LSB. So the display ends up writing the LSB and then may have some time which is dead before they can rewrite the display.